Jump to content

Pancho

Members
  • Content count

    213
  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Community Reputation

12 Good

About Pancho

  • Rank
    Initiate

Personal Information

  • Name
    Tom
  • Location
    Germany

Recent Profile Visitors

2,174 profile views
  1. Quad Sphere ?

    Use a for each loop. In the loop first use a smooth node eg. 10 iteration, mode 1 (important) and then the point normalize wrangle (like in the post before). Repeat the loop somewhere inbetween 10- 100 times and you should get a real nice quad sphere. Made custom nodes for this purpose as this is going to be used so often.
  2. Is AMD potentially risky? (Threadripper)

    What I would like to know is what kind of temperatures other X1950 users achieve with what kind of cooling. Currently there are 4 case fans (200mm) with quite a lot of power, but also with an incredible loud sound. There are other ventilators with about half of the air volume per hour, but much more quiet. Wonder how much the temps will rise, if I exchange them.
  3. Is AMD potentially risky? (Threadripper)

    G.Skill DIMM 128 GB DDR4-3000 Octo-Kit, Arbeitsspeicher(F4-3000C14Q2-128GVRD, Ripjaws V)
  4. Is AMD potentially risky? (Threadripper)

    So far no problems. ASRock X399 Taichi with Air cooling. Temps under load around 56-60°C. 128GB.
  5. You are right. I set it up to shut down the HDDs or mainly everything. But that doesn't make any sense, since the computer is still busy. There should be a trigger to go into hybernation and not some strict time plan. Sometimes I have the feeling that it goes into this mode pretty quickly without 2h of waiting since my last input on the PC. No idea who programmed this part, but it's not working properly. Thanks for yout input!
  6. I like the idea of saving power and in case the PC doesn't do anything for a long period, I'm quite happy if it decides to take some time off. I'm not so happy about my new system entering power saving mode though it is calculating stuff like mad. Wonder why the power saving module doesn't realize that the GPU or CPU is under heavy load. Is there a way to set this up correctly? Never encountered this behaviour before on no other PC of mine. Cheers Tom
  7. All FEM scenes broken in 16.5?

    Thanks, Luke. But sometimes the know how from yesterday needs to incorportated in today's scene. Point is, I need to find an alternative for the stuff I need to do. If targetv doesn't work anymore, than SESI should come up with a page and document why this is so and what workflow replaces the former one. Right now it seems I can save myself with targetP, but not knowing what's going on disables you in recreating your workflow properly.
  8. All FEM scenes broken in 16.5?

    Here's my scene set up. test_001_a_0004.hiplc
  9. All FEM scenes broken in 16.5?

    I try to open several scenes which worked fine before (presumably in 15.5) and now in 16.5 don't. Simple scenes with cloth, FEM and SOP solvers. When I open the scene I get a long list of unrecognized parameters. What annoys me most is the fact that the SOP solver (with enable solver) can set targetv, but the cloth object doesn't care the slightest. FEM solver and SOP solver are both connected to the multisolver. The points do have targetv set, but there is no force applied to the points. v is 0 on all channels? This worked before! What has changed? Since this stuff happens all the time with Houdini, are there any pages on the web which document these changes and what kind of consequences this has on the workflow? But for now: Why is targetv of no interest to the FEM solver? Cheers Tom
  10. Two WEDGE node questions

    Hi Miles, yes. I discovered that you can have multiple entries in the WEDGE node, but the way I took is to have three of them, since I didn't know how the way you propose will evaluate. So far this problem seems to be solved. I'm just rendering. What I did: Three wedge nodes connected to each other. Each wedge node refers to the same ROP. In my heightfield geo I use a null with three custom float inputs. Each of these is connected to the wedge channel of a wedge node. One float entry, one wedge node. My heightfield erode is taking its values from the NULL and by using the pow function I get my 1, 2, 4, 8, ... values and so on. The ROP node uses the referenced values in order to come up with all the different filenames. I'll post some images later on. My only worry now is how to evaluate the results visually. Three wedges with 10 iterations = 1000 variations. : (
  11. Quick and simple: a) How do I combine multiple wedge nodes? There seems to be only one $WEDGE and $WEDGENUM variable. I want to have several settings change, e.g. evaporation rate and rainflow_spreaditerations in the heightfield erode node during the sim. Four iterations through wedge node A and 5 through wedge node B should result in twenty different simulations. But how do I save the files if there is only one $WEDGENUM? In this case I need two wedgenums, in other cases probably more. The range and number of steps seem to be linear. In some cases it is more apropriate to use 1, 2, 4, 8, 16,.... as values. Is this possible? Thanks for any advice! Tom
  12. From what I understand the gas adjust elasticity node is able to change the rest length between the particles. I imagine that by doing so the grain particles can move further apart without breaking the constraint. At least that is what I hope, have the particles further apart but still connected, like something slimy. There have been many request on this subject, but the info given so far is a bit dull. The only tutorial is about gas strain, but I guess that is somehow implemented in grains already. But the rest length changes aren't documented anywhere as far as I can tell. Anybody an idea how to use this node? When I connect it, I always get an exclamation mark and I don't see an effect. So, what is the secret? Cheers Tom
  13. Realistic forces for flip waves

    I'm currently experimenting with forces to create breaking waves in FLIP. I manage to create the general shape of a breaking waves by using a custom vel volume. One problem is that the force still works when the wave is breaking. Obviously the breaking/falling particles shouldn't experience any upwards forces anymore. Of course I can limit forces to only apply if the particles are beneath water level. But I wonder if there is a more realistic approach. A way to determine whether the particles should still experience a force or not. In a way this should be determined by whether or not these particles are "connected" to the main ocean. Another approach could be to check if there are any particles in the direction where the force is coming from. No particles, no way to transfer the vel. But since there are soooo many solvers and ways to deal with FLIPs, is there a general approch to this problem? Cheers Tom
  14. Group by detail not working

    Got it. @id == `detail("../repeat_begin1_metadata1", "iteration", 0) `
  15. I run against a wall. Group expression: @id == detail("../repeat_begin1_metadata1", "iteration", 0) It doesn't evaluate as expected. Getting in the detail attribute this way doesn't work. WHY!? And what is the correct way to do so? Thanks in advance!
×